April, '08] JOURNAL OF ECONOMIC ENTOMOLOGY 125 



extensively studied are found to harbor parasitic species of each type. 

 A similarity in form of body or type of structure is generally seen 

 to exist between the hosts in cases where a parasite attacks several dif- 

 ferent insects. Thus all of the hosts of the aforementioned Ptero- 

 malus are caterpillars, and the same rule obtains among others, for ex- 

 ample, among species of egg-parasites, which always attack insect 

 eggs, although not always those of the same insect. Remarkable ex- 

 ceptions to this are nevertheless to be noted, for instance, the Chalcidid 

 genus Eupelmus {sensu lata), which attacks both the eggs and larvae- 

 of numerous species belonging to six natural orders of insects. 



It is seen, however, that while the hosts of a given species, genus, or 

 larger group are usually of similar form or habitus, that this simi- 

 larity generally depends upon relationship and is not merely acci- 

 dental, for we do not find ordinarily that insects presenting the same 

 general appearance will have the same parasites.- A species of Ich- 

 neumon, I cyaneus Ores., which attacks both saw-fly larvie and cater- 

 pillars, is a notable exception. 



In other words, dissimilarity seems to act more strongly as a deter- 

 rent than similarity does as a persuasive, in regulating parasitism. 



Occurrence in a similar habitat acts very evidently in some cases to 

 widen the range of hosts attacked. This is especially noticeable in the 

 ease of insects producing galls on plants. We must make allowance 

 for the greater care with which the parasites of these insects have 

 been studied, but it is nevertheless astonishing to see what a wide 

 range of species are often regularly attacked by the same parasite. 



The European Chalcid-fly, Ormyrus Uihulosus, has been minutely 

 studied by Mayr, who has bred it from no less than 27 species of Cyni- 

 pid galls, and I have from Massachusetts what is apparently the same 

 species, bred from about half as many North American species by the 

 late Dr. ]\I. T. Thompson. The galls formed by the various hosts of 

 this species are many of them entirely dissimilar in form, the only 

 resemblance between them, aside from their gross gall-like form, being 

 their more or less uniform habitat, attached to twigs and leaves. 



On the other hand, an isolated environment usually restricts greatly 

 the possible parasites of a given species. A case in point is seen with 

 species living beneath bark or boring into wood. INIany groups are 

 represented among the enemies of such insects, but all must come from 

 groups provided with an elongated ovipositor, with which to reach the 

 host for egg deposition. A beautiful case of complete restriction is the 

 Ichneumonid genus Thalessa, which attacks certain wood-boring Siri- 

 cidcB, deposting its eggs in the body of its host far within the infested 

 tree by means of its enormously elongated ovipositor. It may perhaps 



